Systematic Review ajog.org Progestogens in singleton gestations with preterm prelabor rupture of membranes: a systematic review and metaanalysis of randomized controlled trials Johanna Quist-Nelson, MD; Pamela Parker, MD; Neggin Mokhtari, MD; Rossana Di Sarno, MD; Gabriele Saccone, MD; Vincenzo Berghella, MD reterm prelabor rupture of mem- P branes (PROM), defined as <37 OBJECTIVE DATA: Preterm prelabor rupture of membranes occurs in 3% of all preg- weeks gestational age, occurs in nancies. Neonatal benefit is seen in uninfected women who do not deliver immediately approximately 3% of pregnancies and after preterm prelabor rupture of membranes. The purpose of this study was to evaluate contributes to 40% of preterm births.1 whether the administration of progestogens in singleton pregnancies prolongs preg- Neonatal outcomes may be improved nancy after preterm prelabor rupture of membranes. 2 with expectant management in the STUDY: Searches were performed in MEDLINE, OVID, Scopus, EMBASE, ClinicalTrials.gov, absence of infection to facilitate delivery and the Cochrane Central Register of Controlled Trials with the use of a combination of at a later gestational age. Yet, many keywords and text words related to “progesterone,” “progestogen,” “prematurity,” and women with preterm PROM often “preterm premature rupture of membranes” from the inception of the databases until deliver within 1 week.3 Antibiotics safely January 2018. We included all randomized controlled trials of singleton gestations after extend latency and decrease the risks of preterm prelabor rupture of membranes that were randomized to either progestogens or maternal and neonatal infection after control (either placebo or no treatment). Exclusion criteria were trials that included women e preterm PROM.4 6 who had contraindications to expectant management after preterm prelabor rupture of Progestogen administration has been membranes (ie, chorioamnionitis, severe preeclampsia, and nonreassuring fetal status) e studied in different populations7 12 and trials on multiple gestations. We planned to include all progestogens, including but not and has been shown to prolong a limited to 17-a hydroxyprogesterone caproate, and natural progesterone. pregnancy in specific populations that STUDY APPRAISAL AND SYNTHESIS METHODS: The primary outcome was latency from are at risk of prematurity, including randomization to delivery. Metaanalysis was performed with the use of the random effects women with a previous spontaneous model of DerSimonian and Laird to produce relative risk with 95% confidence interval. preterm birth9,12 and those with a Analysis was performed for each mode of progestogen administration separately. 8,11 short cervical length. These thera- RESULTS: Six randomized controlled trials (n¼545 participants) were included. Four of the pies generally are initiated in the sec- included trials assessed the efficacy of 17-a hydroxyprogesterone caproate; 1 trial ond trimester when risk of preterm assessed rectal progestogen, and 1 trial had 3 arms that compared 17-a hydrox- yprogesterone caproate, rectal progestogen, and placebo. The mean gestational age at time From the Department of Obstetrics and randomization was 26.9 weeks in the 17-a hydroxyprogesterone caproate group and 27.3 Gynecology, Division of Maternal-Fetal weeks in the control group. 17-a Hydroxyprogesterone caproate administration was not Medicine, Thomas Jefferson University Hospital, Philadelphia, PA (Drs Quist-Nelson, Parker, and found to prolong the latency period between randomization and delivery (mean difference, Berghella); the Department of Obstetrics, 0.11 days; 95% confidence interval, e3.30 to 3.53). There were no differences in mean Gynecology and Reproductive Sciences at gestational age at delivery, mode of delivery, or maternal or neonatal outcomes between the Magee-Women’s Hospital of University of 2 groups. Similarly, there was no difference in latency for those women who received rectal Pittsburgh Medical Center, Pittsburgh, PA progesterone (mean difference, 4.00 days; 95% confidence interval, e0.72 to 8.72). (Dr Mokhtari); and the Department of Neuroscience, Reproductive Sciences and CONCLUSION: Progestogen administration does not prolong pregnancy in singleton Dentistry, School of Medicine, University of gestations with preterm prelabor rupture of membranes. Naples Federico II, Naples, Italy (Drs Di Sarno Key words: a and Saccone). 17- hydroxyprogesterone caproate, latency, preterm prelabor rupture of Received Feb. 8, 2018; revised March 20, 2018; membranes, progesterone, progestogens, PROM accepted March 24, 2018. The authors report no conflict of interest. Corresponding author: Vincenzo Berghella MD. birth is identified by history or on ul- are thought to have to do with [email protected] trasound examination. reduction in uterine contractility,13 0002-9378/$36.00 14 ª The underlying mechanisms of antimicrobial protein up-regulation, 2018 Elsevier Inc. All rights reserved. fl https://doi.org/10.1016/j.ajog.2018.03.027 how progesterone prolongs pregnancy, immunosuppression, and in ammatory although not completely understood, inhibition.15 SpecifictopretermPROM, MONTH 2018 American Journal of Obstetrics & Gynecology 1 Systematic Review ajog.org progesterone. Any route of administration AJOG at a Glance (eg, oral, intramuscular, rectal, vaginal) Why was this study conducted? was included. The purpose of this study was to investigate whether progestogen administration Before data extraction, the protocol after preterm prelabor rupture of membranes (PROM) increases latency of was registered with PROSPERO (Regis- pregnancy. tration number: CRD42017068717). The metaanalysis was reported accord- Key findings ing to the Preferred Reporting Item for Progestogen administration did not prolong pregnancy in singleton gestations Systematic Reviews and Meta-analyses 19 with preterm PROM. (PRISMA) statement. What does this add to what is known? Risk of bias This study clarifies the evidence that progestogen administration should not be The risk of bias in each included study commenced once a patient has experienced preterm PROM. was assessed by 2 authors (J.Q-N. and P.P.) who used the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions.20 Seven do- progesterone has been shown to related to “progesterone,”“prematurity,” mains related to risk of bias were inhibit the tumor necrosis factor and “progestogens,”“membrane rupture,” assessed in each included trial because thrombin-induced mechanisms of mem- “17 hydroxyprogesterone,”“vaginal pro- there is evidence that these issues are brane weakening.16 Women with preterm gesterone,”“tocolysis,” and “preterm associated with biased estimates of PROM also have been shown to have premature rupture or membranes” from treatment effect: (1) random sequence lower levels of progesterone receptor inception of each database until March generation, (2) allocation concealment, membrane component 1, which is a pro- 2018 (Supplement 1). To locate addi- (3) blinding of participants and tein that is mediated by progesterone to tional publications, we reviewed bibliog- personnel, (4) blinding of outcome stabilize the membrane.17 Although pro- raphies of identified studies and reviews assessment, (5) incomplete outcome gestogens generally are initiated between articles. No restrictions for language or data, (6) selective reporting, and (7) 16 and 20 weeks gestation, initiation of geographic location were applied. other bias. Review authors’ judgments progestogens up to 27 weeks18 is associ- were categorized as “low risk,”“high ated with a decrease in the risk of preterm Study selection risk,” or “unclear risk” of bias.19 birth. Given this benefit in the late second We included all randomized controlled trimester and the mechanisms of preg- trials of singleton gestations after pre- Outcomes nancy prolongation in women with a risk term PROM that were randomized to Data abstraction was completed by 2 of preterm birth, it is reasonable to suspect progestogen vs control (either placebo or independent investigators (J.Q-N., G.S.). that the administration of progestogens no treatment). All published random- Each investigator independently would result in prolonging pregnancy after ized controlled trials on any type of abstracted data from each study and preterm PROM. progestogens after the diagnosis of pre- analyzed data separately. The aim of this systematic review and term PROM at <37 weeks gestation were The primary outcome was time from metaanalysis of randomized controlled reviewed carefully. Exclusion criteria randomization until delivery (ie, trials was to evaluate the therapeutic included women with short cervix, latency). This outcome was chosen benefitinprolongingpregnancybythe quasi-randomized trials (ie, trials in because neonatal outcomes after preterm administration of progesterone therapy in which allocation was done on the basis of PROM are correlated with gestational age singleton gestations after preterm PROM. a pseudo-random sequence [eg, odd/ at delivery.21 Additionally, increased la- even hospital number or date of birth, tency in a patient with preterm PROM Materials and methods alternation]) and trials that included improves survival22 without increasing Search strategy women who had contraindications to incidence of adverse neonatal out- The research protocol was designed a expectant management after preterm comes.2,23 The secondary outcomes were priori, defining methods for searching PROM (ie, chorioamnionitis, severe preterm